If the pins are standard pitch it's very easy to protect these with a blank insulated jumper.
Whilst I do understand the reluctance to have the pins bare on an unboxed RPi the thought that I need to solder something to a new pristine board to access an engineered function abhores me as much as having to by pass built in poly fuses.
It smacks of incomplete engineering if the funcion is there then make it easily accessable the board allready has bare GPIO pins not to mention the lack of any sort of cover when supplied tends to negate a lack of potential insulation on two particular pins.
I know this is not an argument for this thread as strictly speaking it's off topic this is more a debate suited to additions I would like to see made to the RPi.
I suppose it is a suggestion that the Foundation lacked some joined up thinking or lazyness in following through as the alternative is braking and remaking the power train at some other point.
I would like to think that the original idea was to place a small surface mount reset button at this point and some bean counting accountant suggested removing the switch could save coppers allowing the RPi to hit a price point.
I would allways prefer to blame an Accountant over an Engineer.